Transfer learning has been instrumental to many applications in language and vision . Yet, very little is known about its usefulness in 3D point cloud understanding . We see this as an opportunity considering the effort required for annotating data . We hope these findings will encourage more research on unsupervised pretext task design for 3D deep learning . We achieve improvement over recent best results in segmentation and detection across 6 different benchmarks for indoor and outdoor, real and synthetic datasets — demonstrating that the learned representation can generalize across domains. Furthermore, the improvement was similar to supervised pre-training, suggesting that future efforts should favor scaling data collection over more detailed annotation, rather than more detailed data collection, says the authors. We hope this findings will encouraged more research into 3D task designs for 3d deep learning, say the authors . Back to Mail Online Online home. Back to the page you came from: .

Links: PDF - Abstract

Code :


Keywords : d - data - learning - article - news -

Leave a Reply

Your email address will not be published. Required fields are marked *


Enjoy this blog? Please spread the word :)